ABSTRACT The goal of this study is to elucidate the epigenetic mechanisms underlying dynamic genome-wide gene expression during spermatogenesis. The gene expression program of germ cells is distinct from that of somatic lineages. Importantly, the somatic gene expression program is largely suppressed in male germ cells. Instead, male germ cells retain a unique cellular identity that is passed on to sperm and gives rise to a totipotent zygote after fertilization. Our recent RNA-seq analysis showed that about three thousand spermatogenesis-specific genes are activated, while approximately three thousand genes expressed in both somatic lineages and progenitor cells of the male germline (termed somatic/progenitor genes) are largely suppressed during late spermatogenesis, i.e., in meiosis and in postmeiotic spermatids. We identified SCML2 as the suppressor of somatic/progenitor genes. SCML2 is a germline-specific subunit of the Polycomb repressive complex 1 (PRC1), a regulator of heritable gene repression during development. We have discovered that Polycomb complexes determine the gene expression profile by programming genes for both repression and activation. Our combined results suggest that the epigenome of undifferentiated spermatogonia is preset (termed ?preprogrammed?) both for subsequent genome-wide gene silencing and activation during spermatogenesis (termed ?programmed differentiation?). What remain unknown are the mechanisms whereby Polycomb proteins regulate gene expression during spermatogenesis. Our central hypothesis is that Polycomb proteins cooperate to preprogram the epigenome in undifferentiated spermatogonia, thus regulating the subsequent dynamic genome-wide expression profile and programmed differentiation necessary for spermatogenesis. This study will address how the epigenome of undifferentiated spermatogonia is prepared to respond to differentiation cues and, afterwards, how the differentiation program is maintained through mitotic and meiotic divisions. We have designed two complementary specific aims. In Aim 1, we will elucidate how PRC1 defines heritable gene activation and silencing during spermatogenesis. In Aim 2, we will address how SCML2 preprograms the epigenome for later spermatogenic differentiation. These studies will reveal novel epigenetic mechanisms by which interplay between Polycomb proteins regulates the dynamic gene expression during spermatogenesis.